Americans spend billions of dollars annually on the detection and quantification of chemical substances. Most of these measurements are performed in well-outfitted laboratories, requiring skilled personnel, large amounts of costly reagents, and long analysis times. Also, the demands for use in clinical point-of-care testing or for field deployment necessitate small, integrated analytical platforms. Many of these needs have helped to spark chemical sensor development [1]. Similarly, the ever growing need to simultaneously measure “everything” in a sample [2] has pushed the development of artificial “noses” and “tongues” [3] which depend upon chemical and biochemical sensor array strategies [4-10]
Presently, there is a need to develop new devices which overcome the disadvantages of presently used methods. Detection methods which allow the simultaneous quantification of multiple analytes in a sample, are less expensive and more simple to construct and operate, are accurate, precise and reliable, and provide adequate detection limits and selectivity would be a welcome advance in the field of analyte detection.
On general method which has been tried is the use of “biosensors.” In the generic biosensor, an immobilized biorecognition element (e.g., an antibody, aptamer, DNA oligonucleotide, enzyme, lectin, signaling protein, transport protein) serves to selectively recognize a target analyte and the binding or conversion (if the analyte is a substrate) event leads to an optical, mass, thermal, or electrochemical response that is related to the analyte concentration within the sample.
Although biosensor development may appear simple, there are many fundamental issues associated with developing analytically useful biosensors. For example, traditional strategies depend upon identifying an appropriate biorecognition element that can selectively recognize the target analyte. A suitable detection/transduction method is used and the biorecognition element is immobilized [11-13] such that it retains its native activity/affinity and selectivity. The biorecognition element—the biosensor's heart in a traditional design—needs to remain stable over time, the target analyte needs to have access to the biorecognition element, and the analyte-biorecognition element association/interaction needs to be reversible or at least easily dissociated/reset following each measurement. The foregoing shortcomings have limited the application of biosensors in analyte detection.
Over the past decade, the introduction of specific binding domains within synthetic polymers by template-directed cross-linking of functional monomers has attracted considerable attention [15]. Molecular imprinting involves arranging polymerizable functional monomers around a template (pseudo-target analyte or the actual target analyte) followed by polymerization and template removal. Arrangement is typically achieved by: (i) non-covalent interactions (e.g., H-bonds, ion pair interactions) or (ii) reversible covalent interactions. After template removal, these molecularly imprinted polymers (MIP) can recognize and bind specific chemical species (i.e., the template or template analogs).
Potential advantages of MIP-based materials include: specificity comparable to a biorecognition element; robustness and stability under extreme chemical and physical conditions; and an ability to design recognition sites for analytes that lack suitable biorecognition elements. MIPs have been developed for (not an exhaustive list) proteins, amino acid derivatives, sugars and their derivatives, vitamins, nucleotide bases, pesticides, pharmaceuticals, and polycyclic aromatic hydrocarbons. However, according to Lam [16], one of the major issues in the development of MIP based biomimetic sensors is signal transduction.
There are several reports of MIP based sensors that exploit luminescence as the transduction modality. For example, the Powell group [17a] formed cAMP-imprinted organic polymers by using trans-4-[p-(N,N-dimethylamino)stryl]-N-vinylbenzylpyrimidinium chloride (fluorophore), trimethylolpropane trimethacrylate, 2-hydroxyethyl methacrylate, and the initiator, 2,2′-azobisisobutyronitrile (AIBN). These MIPs showed a 20% change in fluorescence in the presence of 1 millimolar cAMP and they were selective for cAMP over cGMP. The Murray group [17b] prepared Soman-imprinted organic polymers by using Eu(R)3(NO3)3 (R=pinacolyl methylphosphonate or divinylmethyl benzoate) (fluorophore), styrene, and AIBN. These MIPs were able to detect Soman down to 750 parts per quadrillion and interferences from organophosphorous pesticides was minimal. The sensor response time was 8 min. The Takeuchi group [17c] reported a fluorescence-based MIP sensor for the detection of 9-ethyladenine (9-EA). This sensor was based on templating 9-EA with 5,10,15-tris(4-isopropylphenyl)-20-(4-metharcyloyloxyphenyl)porphryin zinc (II) (fluorophore) and methacrylic acid. In CH2Cl2, these polymers exhibited a 9-EA binding affinity of 7.5×105 M−1, were selective over adenine, 4-aminopyridine, and 2-aminopyridine, and yielded a fluorescence change of 40% in the presence of 250 micromolar 9-EA. The Wang group [17d] reported on a fluorescence-based MIP sensor for detecting L-tryptophan that used a dansylated dimethylacrylic acid monomer (fluorophore), ethyleneglycol dimethylacrylate, and AIBN. In operation the authors loaded a mobile quencher, 4-nitrobenzaldehyde (4-NB), into the MIP which quenched the dansyl emission. Upon addition of L-tryptophan some of the 4-NB was liberated/blocked from accessing the dansyl residue and the dansyl fluorescence increased. The change in fluorescence upon adding 10 millimolar L-tryptophan was 45%. The presence of an equivalent amount of D-tryptophan, L-phenylalanine, and L-alanine caused 32%, 27%, and <9% changes in fluorescence. The Lam group [16] used a photoinduced electron transfer (PET) strategy to form a fluorescence-based MIP for the detection of 2,4-dichlorophenoxyacetic acid (2,4-D) within a templated sol-gel-derived xerogel. In this work, the authors copolymerized 3-[N,N-bis(9-anthrylmethyl)amino)] propyltriethoxysilane (fluorophore) with tetraethoxysilane (TEOS) and phenyltrimethoxysilane (PtrMES) using 2,4-D as the template. The so formed MIP exhibited a change in fluorescence with pH (apparent pKa near 7.2) and it yielded a 15% decrease in fluorescence in the presence of 750 micromolar 2,4-D. Tests with benzoic acid and acetic acid at similar concentrations did not cause significant interference.
Most recently, Edmiston and coworkers [17e] reported an approach to fabricate a fluorescence-based xerogel MIP for the detection of the pesticide 1,1-bis(4-chlorophenyl)2,2,2-trichloroethane (DDT) by using a sacrificial spacer (SS) scheme [18] wherein they reacted 3-isocyanatopropyltriethyoxysilane with 4,4′-ethylidenebisphenol to form the SS. They then prepared the fluorescent monomer by reacting 3-aminoproplytriethoxysilane (APTES) with the fluorophore 4-chloro-7-nitrobenzofurazan (NBD) (attaching the NBD to the APTES amine, NBD-APTES). The imprinted xerogel was then formed by mixing NBD-APTES, SS, and bis(trimethoxysilyl)benzene followed by a typical acid hydrolysis protocol. Once the xerogel was formed, the authors cleaved the SS carbamate bond with dilute LiAlH4 to form amine residues within the template site, and liberating the SS from the xerogel. The sensor responded to DDT (3% change in NBD fluorescence) and the templated xerogels offered selectivity for DDT over potential interferents (e.g., anthracene (A), 2,2-bis(4-chlorophenyl)-1,1-dichloroethylene (p,p-DDE), 1-(2-chlorophenyl)-1-(4-chlorophenyl)-2,2-dichloroethane(o,p-DDD), 2,2-bis(4-chlorophenyl)-1,1-dichloroethane (p,p-DDD), diphenylmethane (DPM), 4,4′-dibromobiphenyl (DBBP), 4,4′-bis(chloromethyl)-1,1′-biphenyl (BCP)). The DDT detection limits were at the single digit part per billion level.
However, in all previous work on luminescence based MIP sensors, no strategy has been developed to ensure that the luminescent reporter molecule is actually in immediate proximity to the analyte when the analyte binding occurs.